CLUSTERS OF GALAXIES, X-RAY OBSERVATIONS CHRISTINE JONES The luminous material in clusters of galaxies falls primarily into two forms-the visible galaxies and the hot, x-ray-emitting intracluster medium. A visible light image of a cluster shows an overdense region of galaxies. The richest, densest clusters contain predominantly elliptical and lenticular (SO) galaxies, while in less dense clusters up to half the galaxies are spirals. Observations of the relative velocities of galaxies (the velocity dispersion) in rich clusters result in mass-to-light ratios of -250 in solar units-the mass-to-light ratio of the Sun is one. Thus with a mass-to-light ratio of -8 for individual galaxies, only about 3% of the total cluster mass is contained within the visible galaxies. X-ray observations provide a different view from that obtained at visible wavelengths. Although emission from individual galaxies in the cluster is sometimes seen, the primary source of x-ray emission is thermal bremsstrahlung from a hot, intracluster medium (ICM). X-ray Juminosities of clusters range from 10x42-10x45 ergs -1 with gas temperatures of 10x7-10x8 degrees kelvin, comparable to the equivalent temperatures as measured by the velocity dispersions for the galaxies in the cluster. In the cores of rich clusters, the mass of gas is -10% of the total cluster mass. The hot intracluster medium is a major observed luminous component of clusters with a mass equal to or greater than that in the stellar component of the galaxies. Thus, understanding the formation of clusters, and the galaxies within them, requires knowledge of the origin and evolution of the hot intracluster gas. The visible galaxies and the x-ray emitting hot gas do not comprise most of the cluster mass. Instead, most of the mass in rich clusters is "dark matter." Although this material is not directly observable at any wavelength and its nature remains unknown, x-ray and visible light observations are used to determine the amount and distribution of the dark matter. CLUSTER MORPHOLOGY AND STRUCTURE Present epoch clusters display a wide variety of properties both in visible light and at x-ray wavelengths, which can be understood in a framework of cluster evolution (dynamical evolution). Rich clusters with low x-ray luminosities, cool intracluster gas (******** degrees kelvin), and low velocity dispersions have longer dynamical time scales and show more substructure. The very x-ray-luminous clusters with hot gas (10x8 degrees kelvin) and high velocity dispersions have shorter dynamical time scales and are relaxed systems. One property that does not fit neatly into this scenario is the presence of a massive, centrally located galaxy. It had been suggested that the importance of a central galaxy was directly related to the evolutionary stage of a cluster. However, x-ray observations show that there is a class of clusters whose properties are those of dynamically young systems which nevertheless contain a massive, centrally located galaxy. More recent theoretical studies of cluster dynamics argue that the importance of the central galaxy is determined at early stages of cluster evolution. Thus, the suggestion has been made that a second parameter-the importance of a central galaxy-be added to the dynamical time scale to generate a two-dimensional cluster classification system as described in Table 1 and Figure 1. Figure 1 illustrates the classification system by comparing cluster optical images with their x-ray emission shown as isointensity contours. The clusters on the top (A1367-left and A262-right) are examples of unevolved systems with long dynamical time scales. The clusters shown at the bottom (A2256-left and A85-right) are dynamically more evolved systems. The clusters on the left do not have a central, dominant galaxy, whereas those at the right have a bright galaxy around which the x-ray emission is clearly centered and concentrated. The clusters with bright galaxies at the centers of their x-ray emission comprise nearly 40% of those clusters surveyed with Einstein satellite observations, which, in this sample, is twice the incidence of clusters selected optically as having a central, dominant galaxy. Related to the morphology of clusters is the presence of substructure. X-ray images of the emission from the intracluster gas are particularly effective for determining cluster structure because the gas effectively maps the distribution of the total underlying mass. Based on an x-ray imaging survey of about 150 rich clusters with the Einstein observatory, substructure is a common phenomenon. About 30% of the clusters have multiple peaks, of which two-thirds are double and one-third are more complex. The fraction of clusters showing substructure in their x-ray emission agrees with that determined from optical studies of galaxy distributions. These multiple-peaked structures are evidence for a still-evolving cluster potential. Figure 2 shows two examples of double and complex systems. The remaining 70% of clusters show single peaks in their x-ray surface brightness distributions. Of these clusters, roughly half have bright galaxies at their x-ray centers. These clusters tend to have high gas densities in the central regions, so the gas can radiate its energy and cool on relatively short (billion-year) time scales. Of those clusters with single peaks, about 40% have cooling times less than the age of the universe. For half of these, the estimated rates at which the gas cools and flows toward the cluster center exceed 100 ******. Although often referred to as "cooling flows," little matter is flowing and the flow rate is slow (velocities of tens of kilometers per second), so that the gas generally remains in a near-equilibrium state. Perhaps the best-studied example of a cooling flow cluster is the Perseus cluster. Observed temperatures for gas in the Perseus cluster range from the cluster mean (***** degrees Kelvin) to a factor of 10 lower. The observation of gas with temperatures below the cluster mean is the strongest evidence for mass deposition. For the Perseus cluster, estimates of the cooling rates for the gas at each observed temperature yield consistent mass deposition rates of **************. For the different temperature regimes, the cooling times of the gas span a range up to several billion years, suggesting that cooling flows are long lived. Detailed analysis of x-ray observations of the Perseus cluster shows that the gas is not all flowing to the center of the cluster, but is cooling out over a wide range of radii. Although high mass deposition rates can contribute significantly to the mass of the central galaxy, the precise fate of the cooling gas remains uncertain. Optical observations of the central galaxy show that any new star formation can produce only a very few stars with masses exceeding one solar mass. The large mass deposition rates inferred for many clusters have been controversial. A variety of suggestions have been made to attempt to reduce the calculated mass deposition rates to values which would not significantly affect the mass balance of the central galaxies. Models to reheat the cooling gas through thermal conduction from the outer, hotter gas, or through supernovae, galaxy motions, and relativistic particles have not been successful in reducing the inferred cooling rates. In conclusion, either cooling flows deposit a significant amount of mass around central cluster galaxies, or curtailing a large cooling flow requires a heat source capable of providing energy up to **** ergs **** for a total of **** erg over the lifetime of a cluster. CLUSTER MASS DETERMINATIONS The x-ray emitting gas, while itself an important mass component of clusters, also can be used to trace the total underlying mass, which determines the gravitational potential. The intracluster medium relaxes on a relatively short time scale to hydrostatic equilibrium, which means that any flow velocities in the gas are small. Thus the equation for a hydrostatic gas can be used to determine the cluster gravitational mass. This application of the x-ray observations requires the determination of both the gas density, which can be determined from the observed x-ray surface brightness, and the temperature distributions. Gas temperature distributions have been obtained for only a few systems and indicate the presence of dark matter in agreement with optical determinations. Future x-ray satellite observations will provide spatially resolved measurements of the gas temperature for many clusters and will allow the accurate determination of cluster mass distributions. THE ORIGIN OF THE INTRACLUSTER MEDIUM In addition to providing information on the structure and morphology of clusters and the total mass distribution, the study of the intracluster medium is important in its own right. Because the ICM mass is equal to or greater than that in stars, it is of particular importance to determine the origin of this large fraction of the known baryonic mass. One of the most important results related to the origin of the ICM was the discovery of emission lines from iron and other heavy elements in the energy spectrum of the hot gas. Cluster x-ray spectra show both that the x-ray emission is thermal in origin and that the heavy element abundances are between 20 and 50% of the solar values. Because heavy elements can be produced only through thermonuclear reactions in stars or by supernovae, the discovery that the intracluster medium was rich in heavy elements requires that material processed through stars be ejected into the ICM. While the enriched material must come from the galaxies (in the absence of a very early stellar population outside of galaxies), recent studies have suggested that the bulk of the ICM of a rich cluster could not have originated within the galaxies because the ICM mass is up to several times larger than the mass of the galactic stellar component. Thus, a considerable fraction of the ICM in rich clusters must be left over from the formation of galaxies. FUTURE OBSERVATIONS This entry has highlighted some of the properties of clusters of galaxies in which x-ray observations have been particularly useful. However, the study of x-ray emission from clusters has only begun. Other important contributions will arise as more powerful observatories become available. Detailed x-ray studies of distant clusters, combined with radio observations, have the potential for determining a precise value of the scale of the universe, the Hubble constant, whose value has eluded astronomers for decades. Measurements of the abundance of heavy elements in the intracluster medium for clusters at different look-back times will provide information on the formation of galaxies and nucleosynthesis in stars. Mass determinations in clusters may elucidate the nature of the dark matter. Through the unique view of clusters of galaxies provided by x-ray observations, we can determine fundamental properties of galaxies, clusters, and the universe. Additional Reading Fabian, A.(1988). Cooling Flows in Clusters of Galaxies. Kluwer Academic, Dordrecht. Forman, W. and Jones, C.(1982). X-ray-imaging observations of clusters of galaxies. Ann. Rev. Astron. Ap. 20 547. Sarazin, C.L.(1986). X-ray emission from clusters of galaxies. Rev. Mod. Phys. 58 1. See also Galaxies, X-Ray Emission; Intracluster Medium.